Data Warehouse Hybrid Modeling Methodology

Abstract

The classic conceptual modeling around business processes followed by the ‘bus matrix’ methodology of designing the data cubes of data warehouses (Kimball & Ross 2013). For a serious system, such a quantity of management questions and dimensions, the bus matrix results a difficult-to-understand conceptual data model. The subject of automation and conceptual design – to which many individual methods already have been developed – are relevant topics in today’s literature also. In the 2010s data warehouse projects were realized in Hungarian higher education to inform the decision makers of the universities about their own institutions. As we participated in this project in 2009–2010, we faced that our bus matrix at the end contained about 80–120 indicators with nearly 200 dimensions (dimensional attributes), therefore we worked on the early stenography to formalize the management question. We provide a kind of ‘business intelligence problem solving thinking’ and a kind of descriptive language that can serve it and present a method which has two novelties compared to formers: It is based on the management questions and its visualization. As a kind of stenography, it is always based on the terminology corresponding to the current problem, so it forms an intermediate language for the data model. We introduce our method through an example in a popular research area which is activity tracking.

Keywords

• formal methods
• data science
• information management